Candidate allele & gene hypotheses developed from pharmacological & genetic findings are tested. Also, we follow up on whole genome linkage findings. Candidate gene linkage investigations included: DRD2 Ser311Cys and haplotype/alcoholism and opioid addiction, HTR1B/antisocial alcoholism, SLCA4/anxiety, HTT and severe serotonin-associated behaviors, BDNF Val66Met and episodic memory & COMT/opioid addiction/Schizophrenia/anxiety/dysphoria/Executive cognition. Phenotype, sampling framework, genetic structures [e.g. TDT, sibpairs], & power are evaluated. Assessment is generally with structured interview, & usually with intermediate phenotypes and/or additional end-phenotypes. The Finnish dataset was ascertained from criminal alcoholic probands & is thus enriched for Type II [early onset] alcoholism. SW Indian, Plains Indian, Ten Tribes, & Finnish datasets are derived from isolates; psychiatric interviewed controls are available from source populations. Chinese and German case/control datasets are in use. A predominantly African American cocaine dataset [N=800] is in use. Remaining datasets are almost entirely Caucasian. Some 980 psychiatrically interviewed Caucasian population controls are available. The SW Indian dataset includes 600 sibpairs, the Finnish dataset used in the SLCA4 study included 366 sibpairs, & the Plains Indian EEG dataset contains a large number of sib & relative pairs. A pain threshold study [Dionne, Iadarola] is designed as sib-parent quad enabling quantitative sibpair linkage & TDT. Both schizophrenia datasets [Egan & Weinberger, Malhotra] enable TDT/sTDT analysis. Power for case/control association is dependent on allele frequency, association strength, & desired level of alpha. The functional alleles we test have moderate [HTR2C, HTR2A, mu opioid receptor: 0.1] or high [SLCA4 & COMT: 0.4] frequencies, & the functional polymorphisms known to affect alcoholism ADH2 Arg47His & ALDH2 Glu487Lys have 5-10 fold effects. Therefore, our smallest datasets [n=50] have 58% power at the p less than 0.01 level; power is essentially 100% for larger datasets available for alcoholism [several datasets], schizophrenia, OCD, anorexia nervosa, SAD, & bipolar. For lower effect size, power at p less than 0.01 remains high [greater than 0.8] for the association strength of 0.1 in the larger datasets. Relationship of candidate genes and alleles to behavior: DRD2. The DRD2 dopamine receptor, "Reward Deficiency Gene" hypothesis, was tested in SW Indians with three DRD2 markers: the Taq1A marker previously implicated, an STR, & Ser311Cys, which impairs signal transduction & which is far more abundant [0.16] in this particular population as compared to Caucasians [0.03]. Because it impairs function, Cys311 is a surrogate for unknown DRD2 alleles that attenuate function to a similar extent or by the same mechanism [transduction]. There was no sibpair linkage nor association, although the dataset included 15 Cys311/Cys311 homozygotes. Previously, LNG reported the early negative results for DRD2 & alcoholism in ethnically matched, psychiatrically interviewed alcoholics & controls, in severe alcoholics, & in alcoholics with parental alcoholism. LNG identified large ethnic differences in DRD2 allele & haplotype frequencies, & helped discover Ser311Cys. Subsequently, the COGA sibpair linkage study [Edenburg et al] has also failed to detect a DRD2 linkage signal. These results suggest there is scant evidence for a substantial role for DRD2 variation in alcohol vulnerability. To monitor functional variants at DRD2 locus, LNG has developed high throughput assays for Ser311Cys & -141DelT. In work with a ten-locis DRD2 haplotype, we find positive linkage to opioid dependence in two large datasets. Candidate alleles to behavior: Serotonin transporter. A functional serotonin transporter promoter polymorphism was associated with dimensionally measured anxiety. Quantitative linkage in 366 Finnish sibpairs linked this polymorphism to the two anxiety-related subscales of the TPQ [Mazzanti]. Function was further pursued by in vivo B-CIT SPECT imaging of transporter density. In controls [but not alcoholics], the lower transcribing s allele was associated with lower brain transporter density. Alcoholics experience sustained changes in serotonin transporter function, for example due to alcohol-induced serotonin release or effects of withdrawal. The polymorphism was linked to metabolic activity in amygdala induced by a cognitive fear challenge. A rare SERT missense variant was linked to severe psychopathology in two families. We built on findings from mouse genetic models to identify a potential role for the 5HT1B receptor in a subtype of alcoholism. HTR1B was implicated as an alcohol preference gene by a mouse alcohol preference QTL & subsequent discovery that the HTR1B knockout mouse exhibited increased aggression & preference for alcohol. As a terminal autoreceptor, 5HT1B modulates serotonin release & was an excellent candidate for variation in alcohol preference & aggressivity, independently of the mouse findings. Aggressive & impulsive behaviors are represented in two psychiatric diagnoses - ASPD & IED [Intermittent Explosive Disorder] & alcohol preference is a component of alcoholism. For HTR1B, we used G861C [described by us] & the closely linked D6S284 STR. We studied 640 Finnish subjects included 350 sibpairs [220 unaffected, 79 discordant & 51 affected] & 418 Southwestern Indians including 305 sibpairs [223 unaffected, 71 discordant & 11 affected pairs] & classified for the presence/absence of antisocial alcoholism [DSMIII-R alcoholism plus either ASPD or IED]. In this single-locus study, evidence for i.b.d. linkage was found in both datasets [p=0.04 & p=0.01] & association to HTR1B was also detected in Finns. Relationship of candidate genes and alleles to behavior: COMT (catechol-O-methyltransferase). COMT Val158Met & Frontal lobe function: Intermediate phenotype. Executive cognitive functions (ECF)localized to the frontal lobe are thought to be impaired in several psychiatric diseases: alcoholism, ADHD, & schizophrenia SZ). Patients with SZ, & to a lesser extent their healthy siblings, have deficits in working memory & ECFs & show excess cortical activity with fMRI [& are thus said to be inefficient] during these tasks. We found a gene-dosage relationship of a functional COMT variant to frontal cognitive function. The relationship was seen across several populations: SZ, moderate-severe head injury, & controls, which differ substantially in baseline function. Dopamine enhances prefrontal cortical efficiency during frontal lobe tasks. Val158Met, a common COMT variant, leads to four-fold reduction in enzyme activity & was thus an excellent candidate gene for variation in frontal lobe cognitive function. Also, three TDT linkage studies had detected evidence for Val158Met in SZ & some evidence for a SZ locus had been found near 22q11. Wisconsin Card Sort performance was evaluated versus COMT genotype in 75 controls, 184 SZs, & 222 siblings of SZs, with the result that a remarkable allele-dosage relationship was found to preserve errors in both the SZ patients & the controls. This finding was replicated in normal controls and head-injured patients, and then expanded by a study of frontal lobe metabolic activity in individuals evaluated using blood oxygen level dependent [BOLD] fMRI during the N-back task, which accesses prefrontal cognitive functions. The Val158 allele was associated with increased metabolic activity in frontal lobe consistent with the hypothesis of diminished cortical efficiency. The Val allele may have counteradvantages: in a two populations the Met allele predicted anxiety in women and decreased frontal EEG coherence. Met was associated with diminished pain tolerance and ability to activate endogenous opioid activity after a pain/stress challenge.